. . .This only is certain, that there is nothing certain; and nothing more miserable and yet more arrogant than man. — Pliny the Elder

Twentieth Century Inventions - Charles Gibson

Telegraphic Inventions


Telegraphing Photographs

Of the recent inventions in connection with electrical transmission there is no doubt that the telegraphing of photographs is of most general interest. There are now several methods of transmitting photographs electrically, but the most outstanding and the most interesting from the invention point of view is that devised in 1904 by Professor Arthur Korn, then of Munich, now of Berlin.

The general idea is, of course, to cause a photograph to control an electric current, which when transmitted to a distance may rebuild a duplicate of the controlling photograph.

It had been known for a generation that the non-metallic element selenium, when in a two extremes of no light (opaqueness) and full light (transparency) there would be in every photograph an endless variety of light and shade.

In order to bring each part of the photograph beneath the pencil of light, the cylinder is caused to rotate and to move longitudinally, so that the point of light traces a continuous spiral around the cylindrical photograph from end to end. We have the same relationship between a phonograph cylinder and its stylus.

As the pencil of light falls upon the different parts of the photograph in turn, there will be an endless variety in the intensity of the beam of light issuing from the end of the cylinder and falling upon the selenium cell, and there will be a corresponding variation in the electric current passing through the cell. We might picture this ever-varying current passing out on to the long-distance line, but the action of the selenium cell is not quick enough. This led Korn to invent a compensating arrangement, by which he gets rid of this inertia trouble.

The compensation is brought about by using two selenium cells on opposite sides of crystalline condition, had its electrical resistance altered very materially by light falling upon it. Selenium cells are made so that in the dark they prevent the passage of an electric current through them, but when exposed to light they have their electrical resistance so greatly reduced that an electric current passes freely. Between these two extremes of no light and full light, the selenium cell acts according to the intensity of the light which impinges upon it. Professor Korn has made use of such selenium cells in his invention.

Having made a flexible transparent positive of the photograph, the inventor wrapped this around a cylinder of glass. He then caused the light from a Nernst lamp to pass through lenses, which caused a concentrated beam or fine pencil of light to fall upon the transparent photograph. If this spot of light fell upon an opaque part of the photograph, no light would enter the cylinder, but if it fell upon a transparent part the light would pass into the glass cylinder and there fall upon a right-angled prism, which would reflect it out at the end of the cylinder to fall upon a selenium cell. Between these a Wheatstone bridge, and dividing the battery into two parts for the other two sides. Tip first or main selenium cell is so composed that it has small inertia and high resistance. This means that while it will act quickly, it will only let a small amount of electric current pass. The second cell has large inertia but low electrical resistance. The current from the first cell, which is controlled by the photograph as already described, is now shunted across a circuit containing a string galvanometer. In this instrument, there are two fine silver wires carrying a light shutter of aluminum foil, and these are free to move laterally in a strong magnetic field. While no current passes, the little shutter blocks the way of a pencil of light from a second Nernst lamp, and prevents it reaching the second selenium cell. When the wires and shutter are deflected by the current from the first selenium cell, they allow the beam of light to pass, and it is reflected by a prism, which causes it to fall upon the second cell. The greater the current, the greater the deflection of the silver wires, and the greater the action on the second selenium cell.

It will be observed that both cells are working in unison under the control of the photograph, and these constitute two sides of a Wheatstone bridge, which provides a much more sensitive apparatus. Instead of the electric current rising gradually and falling slowly, when the light is suddenly admitted and withdrawn, it rises very rapidly and falls to zero almost immediately.

This controlled current is sent on to the long-distance line, and at the receiving end the current actuates a string galvanometer similar to that used as a compensator in the transmitter. In the receiver, the little aluminum-foil shutter, under the influence of the incoming current, controls a pencil of light from a Nernst lamp. We might picture this ever-varying pencil of light falling upon a sensitive film, wrapped around a rotating cylinder, and thus building up a duplicate of the transmitting photograph. But this first galvanometer is used as a relay to control a local battery current. Instead of falling upon the photographic film the light falls upon a selenium cell, which controls the local battery and operates a second string galvanometer. The aluminum-foil shutter of this instrument controls the light from a second Nernst lamp, and this pencil of light is focused upon a small aperture in a screen which protects the photographic film. When at rest, the shadow of the little aluminum-foil shutter falls upon this aperture, and no light can reach the sensitized film, but when the galvanometer is actuated the shutter is deflected and allows light to pass; the greater the incoming current the greater the intensity of the light passing through the aperture.

It will be obvious that if the rate of longitudinal movement given to the rotating cylinder be great, the resulting picture at the receiving end will be made up of distinct lines, for the pencil of light will trace a comparatively coarse spiral around the receiving drum. On the other hand, if the longitudinal movement be very slow the lines may be made practically to touch one another, and produce a solid effect. No matter how solid the picture may be made to appear, it will in reality be built up of lines varying in thickness according to the movements of the controlling shutter in the string galvanometer. Of course, when the film is taken from the drum and the developed photograph laid flat, there is no appearance of a spiral, but what appears to be a series of fine parallel lines.

One of the many difficulties which Professor horn had to overcome was the fact that the movements of the string galvanometer would not be exactly proportional to the strength of the electric current received. Suppose the galvanometer to be so adjusted that the maximum current which could arrive would deflect the shutter so as to completely uncover the aperture, and allow the maximum of light to pass through the lens to the sensitized film. With half the maximum current, the lens would not be exactly half uncovered, and so on. One method by which the inventor overcame this difficulty was to make the aperture of some form that would admit an increasingly greater proportion of light the further it was uncovered. It was found that a right-angled triangle was the best form of opening. Even this was imperfect, so a piece of toned film, dark at one part and shaded off to transparency, was placed behind the triangular aperture. This was so placed that the point of the triangle which is first uncovered is protected by the dark part of this film, while the shaded part admits an increasingly greater amount of light as the shutter opens wider. By means of this screen placed in front of the lens, the light is made proportional to the electric current received, and also to correspond with the intensity of light at the transmitting station.

At each station there is a transmitter and a receiver placed side by side, but it is found unnecessary to duplicate the string galvanometer arrangement. The one set of galvanometers will serve both for transmitting and receiving.

Another difficulty which the inventor had to overcome was to ensure that the receiving cylinder carrying the sensitized photographic film would keep in exact step with the transmitting cylinder upon which the sending photograph was fixed. Unless there is complete synchrony between these two far-distant cylinders, the result cannot be of any service; the parallel lines would not be placed in proper relationship to one another. A most ingenious method is adopted, whereby the cylinders are synchronized at the end of every revolution. Each cylinder is driven by a similar electric motor, but the motor of the receiving cylinder is set to run about one percent faster than the motor of the distant transmitting cylinder. This is done so that the receiving cylinder will come to the end of its revolution very slightly in advance of the transmitting cylinder. As soon as the receiving cylinder has completed its revolution it is automatically stopped by a projection on a disc on the cylinder shaft being caught by a pawl. The motor still runs on at even pace, but the cylinder, being driven by a friction clutch, is held stationary so long as this pawl remains engaged with the disc on the cylinder shaft. But as soon as the distant transmitting cylinder (which is one percent behind the receiving cylinder) has completed its revolution, it makes an electrical contact, which sends a current along the line wire to an electro-magnet which immediately releases the pawl and allows the receiving cylinder to set off on another complete revolution. And so on the two cylinders go, the receiving one always trying to get in advance, but being held up at the end of each revolution long enough to set off again at exactly the same moment as the sending cylinder. In this way the two cylinders rotate in complete synchrony.

It will be observed that the synchronizing impulse is sent over the line wire, and yet it must not enter the sensitive string galvanometer at the receiving station, for if it did so it would damage the instrument. Protection is secured automatically, for when the receiving cylinder is stopped by the pawl, it cuts out the photographic connections so long as the cylinder remains at rest. As soon as the pawl is released these electrical connections are re-established. This momentary halting of the cylinder, with its accompanying withdrawal of the light, is arranged to take place while the join of the photograph is opposite the beam of light, so that there is no interference with the continuity of the lines producing the picture.

The electric motors which drive the cylinders are run at a speed of about 3000 revolutions per minute, and this high speed is reduced by gearing so that the cylinders revolve thirty times per minute. Before commencing a transmission the operator sets the speed, which is controlled by regulating a resistance which is placed in series with the field magnets of the motor. The operator starts the motor and watches a speed indicator or 1frequency meter1 connected to the motor. This indicator is very ingenious. It consists of an electro-magnet with a laminated core, so that it can respond quickly to changes of magnetism. It is supplied with an alternating electric current from collecting rings fixed on the armature of the motor. The magnetism of the laminated core of the electro-magnet will therefore change its polarity at each half-revolution of the motor, which will be 6000 times per minute. A piece of steel spring, fixed at one end and free to vibrate at the other, is placed over the pole of this electro-magnet, and the steel tongue is tuned to vibrate freely 6000 times per minute (100 times per second). When the magnet is receiving too many or too few impulses from the rotating motor, the vibrating tongue will not respond, but as soon as the speed is 3000 revolutions (giving 6000 impulses) the steel tongue will sing out a clear musical note.

In order that the operator may have a little latitude in the tuning of his motor, the indicator is supplied with other two vibrating steel tongues. One of these is set to vibrate at 99 vibrations per second, and this will be set in motion when the motor is running at 2970 revolutions per minute (99 x 60 2). The other vibrator responds to 101 vibrations per second, and will be actuated by the electro-magnet when the motor is running at 8030 revolutions per minute (101 x 60 2).

While the operators are adjusting the speed of the motors by means of these frequency meters, they use a telephone on the line wire, and mutually adjust their motors to whichever frequency they find convenient. The telephone enables them also to adjust the string galvanometers and to arrange the whole apparatus generally.

As already stated, this system of telegraphing photographs was invented by Professor Korn in 1904. It was used between Paris and London, for the Daily Mirror, from the autumn of 1907 till the summer of 1909, when a different system was installed. This is known as Korn's telautographic system, and one form of it is known as Thorne Baker's telectrographic process. While Professor Korn has worked out this more mechanical system, he has not neglected his selenium process, and he is hopeful of applying it to long-distance submarine cables. He hopes to be the first to telegraph a photograph from Europe to America, and he is aiming at doing so in time for the San Francisco Exhibition (1915).

The telautographic system is not of the same scientific interest as the selenium process, but it is of considerable practical interest. The operation of light plays no part in the telautographic transmission. The photograph which is to be transmitted is printed as a half-tone picture upon a sheet of lead-foil. The printing material is not ink but fish-glue, which is a non-conductor of electricity. In the half-tone reproduction the photograph is broken up into lines, and is wrapped around a metal cylinder, while the point of a metal stylus passes over the revolving picture, just as the spot of light did in the selenium process. An electric current passes through the stylus and the lead-foil to the metal cylinder and thence to the line wire, by which it reaches the distant station. When any part of the fish-glue printing comes beneath the stylus, the electric current cannot pass to the line wire. While a constant electric current is supplied to the stylus, it is an interrupted current which passes out to the distant station. These interruptions in the electric current will correspond with the number and width of the printed lines over which the metal stylus passes.

At the receiving station this interrupted electric current passes to a metal stylus and metal cylinder similar to that already described; but in this case it is a chemically-prepared paper which intervenes between the stylus and the metal cylinder. This paper is absorbent, and the incoming electric current can pass through it. The current produces an electrolytic effect which darkens the paper. If a constant electric current were received, there would be a continuous darkening of the paper, but the lines of fish-glue on the transmitting cylinder interrupt the constant current, so that during these interruptions the receiving paper remains white. In this way a black-and-white duplicate of the fish-glue picture is reproduced at the receiving station. The results are wonderfully good, and this simplified apparatus is at work between Paris and Berlin, and Monte Carlo and Paris, while similar apparatus is used between Paris and London, and London and Manchester. The author has received from Professor Korn a telautographic photograph which was transmitted from Monte Carlo to Paris, a distance of 750 miles, and which took only twelve minutes in transmission.

It is even possible to send a photograph by wireless telegraphy, although this has been done only experimentally at present. In the telautographic system, just described, the electric current under the control of the photograph may operate a wireless transmitter instead of passing out on to the line wire.

At the transmitting station the interrupted current actuates an electro-magnet which attracts a soft iron diaphragm at each impulse. When this diaphragm is attracted and let go, it closes and opens the primary circuit of a transformer, causing sparks to take place at the air-gap in the secondary circuit. This secondary circuit is connected to an aerial and to the earth as in ordinary wireless telegraphy. At each make and break made by the diaphragm, electric impulses are set up in the aerial, and these send out electromagnetic waves through the ether of space to the distant station. These waves are entrapped by an antenna as in ordinary wireless work, and the electric impulses produced by these waves reaching this second aerial, are conducted to a detector, which acts as a very sensitive relay, permitting a local current to pass at each impulse received. This local current passes through the stylus, prepared paper, and the metal cylinder, and the picture is built up in the manner already described. But as time is required for the makes and breaks made by the detector, it is only possible to transmit very simple sketches or diagrams at present. However, the wireless transmission of a plan, showing the disposition of an army, might be very useful in time of war.

Having no line wire, the synchronizing of the cylinders is more difficult in wireless transmission. One method of keeping the revolving cylinders in step with each other is to set them both to travel slightly faster than required. At the end of each revolution they are caught by a pawl and held till the time set aside for each revolution has passed. The release is brought about by a chronometer controlling each cylinder. If the two chronometers keep in exact step with one another then the two cylinders will be perfectly synchronized.

Another method is to adopt a plan similar to that already described in connection with the line system. The receiving cylinder being in advance, throws itself out of action at the end of each revolution, switching off the detector from the photographic apparatus. At the same moment the coherer has switched itself on to a relay circuit, so arranged that when the distant transmitting cylinder sends out the next wireless impulse on the completion of one complete revolution, the relay in the receiver releases that cylinder, and the two cylinders set off in unison once more. This method can be used only over short distances, but the chronometer method gives very fair results and is independent of the distance. The speed of the rotating cylinders is considerably lower than with transmission over wires.

Professor Korn has made some wireless experiments by means of an improved system. In these he used the telautographic method of transmission, but with a selenium cell and string galvanometer in the receiver. The receiving circuit in which the wireless detector is placed is tuned to the distant transmitting circuit, and a constant stream of ether impulses is maintained, to be broken only by the transmitting stylus coming on to a conducting part of the metal-foil. This short-circuits part of the inductance coil and thus alters the rate of the ether impulses, putting them out of tune with the receiver. Therefore the detector in the distant receiving station is actuated only so long as the transmitting stylus is passing over the fish-glue parts of the sending photograph.

The string galvanometer, which is actuated by the detector, is somewhat simpler than that used in transmission with wires. In wireless work the galvanometer consists of a single wire or fine metal thread through which the currents are passed. This conductor will, of course, possess a magnetic field so long as any current passes through it. It will therefore be deflected by an electro-magnet between the poles of which it is stretched. The light from a Nernst lamp is concentrated upon this wire, and an image of the wire is projected on the slit of a tube leading into the box in which the sensitized film is rotated. When the galvanometer is traversed by a current, the wire is deflected and the shadow moves away from the slit, and light reaches the photographic film.

Although wireless photo-telegraphy is only in an experimental stage, there is reason to hope that those experiments will lead on to some practical system which should be of great service.

Page-Printing Telegraph

In 1905, Donald Murray, an Australian journalist, read a paper before the Institution of Electrical Engineers (London), on the subject of setting type by telegraph. About five years earlier he had arrived in New York with a new invention which was to print telegrams in page form, and was to be operated over a single telegraph line.

The idea itself was not new, there having been patented no less than 150 such inventions during the nineteenth century, but Murray's invention was on an entirely different principle. He was not going to attempt to control a complex typewriting machine by means of a single telegraph line; his idea, although not developed fully at that time, was to prepare a perforated paper-ribbon, such as is done for the Wheatstone automatic transmitter. He then proposed to use this paper-ribbon to control an electric current passing out to the telegraph line, and at the receiving station an instrument was to prepare a duplicate ribbon. Then this telegraphically-prepared ribbon was to be run through an automatic typewriting machine, which would print the telegram in page form.

But the first Murray printing machine was primitive, and was described in New York as "a sort of cross between a sewing-machine and a barrel-organ," the operator having to turn a handle to drive the machine. It received nicknames such as "Murray's coffee-mill," the "Australian sausage-machine," and, perhaps more reasonably, the "Baby," indicating that the apparatus required a good deal of nursing.

Printing telegraphs had been used on the Continent for a generation, but these printed the message on a long tape, and their speed was limited to hand signalling, whereas the ?Murray telegraph was to print in page form at a very high speed.

The invention in its practical form consists of four separate machines. First, there must be a perforator for punching the symbols on the paper-ribbon. Then a transmitter for signalling to the distant station. Thirdly, a recorder for receiving the transmitted signals, and preparing a duplicate of the transmitting ribbon. And, finally, a typewriting machine which can be controlled by the perforated ribbon.

The 'perforator' has a typewriter key-board, which is manipulated in the usual manner, but instead of printing Roman characters on a sheet of paper, it punches holes in a paper-ribbon. Just as in the case of the Morse Telegraph, definite combinations of dots and dashes represent all the letters of the alphabet, so in the Murray Telegraph, definite combinations of perforations make up the whole alphabet. There are five punches in the perforator, and when the letter 'A' key is depressed on the keyboard, the first two punches act simultaneously. When the key 'E' is depressed, the first punch only is called into play, while the letter 'T' operates only the fifth punch, and so on until the whole alphabet is completed by different combinations of these five punches. There is a central row of feed-holes, which is punched beforehand. This is done at a great speed by pulling the ribbon through between a punch-wheel and a die-wheel. If desired, it could be done automatically on a large scale, as is the case in perforating sheets of postage stamps.

It will be understood that the speed of operation of the perforator does not necessarily determine the speed at which the messages are to travel over the telegraph line. We know that in the case of the Wheatstone automatic sender, the speed of preparing the perforated ribbon is slow, the punching of each individual signal having to be controlled separately, four movements of the operator's hand being required to signal many of the letters. But the speed of the punching machine bears no relationship to the speed at which the prepared tape may be rushed through the automatic transmitter, while the automatic recorder acts in sympathy. The Murray perforator has the advantage of only one movement of the operator's hand to punch the complete signals forming each letter.

In ordinary typewriting the typist has to go carefully to avoid mistakes, as corrections disfigure the production. But in operating the typewriter keyboard of the Murray perforator the operator can use more freedom. If a mistake is made it does not require to appear in the printed telegram. This advantage belongs in some measure also to the Wheatstone automatic transmitter, in which a mistaken perforation may be cancelled by punching a number of 'A' s,' and then proceeding with the correct perforation. The distant recorder will duplicate these obliterating signals, in the form of a series of dots on the receiving ribbon, but the subscriber knows to take no notice of them in writing out the message. In the Murray telegraph, the obliterating holes appear in the duplicate ribbon produced by the receiving instrument; but when this is run through the printer, that machine stops momentarily till the obliterating holes have passed. Not only does the machine cease printing, but it does not feed the telegraph-form forward until the printing recommences. In this way no trace of the correction appears in the finished telegram.

The perforator is operated electrically, and when a letter-key is depressed it switches on the current to an electro-magnet, which operates the punches on the front stroke of its armature, and on its back stroke feeds the paper-ribbon forward one letter-space. It will be obvious that the operation of the perforator is not dependent upon any knowledge of the signals on the part of the operator; the manipulation is the same as in an ordinary typewriter.

The prepared ribbon is then used to control the electric current passing out to the distant station. As already stated, the object of the Murray single-line transmitter is to reproduce, at the receiving station, a duplicate of the perforated ribbon. There are two well-known methods of automatic transmission from a perforated ribbon. In the one system the electrical contact is made directly through the holes in the ribbon. So long as the solid ribbon intervenes between the rolling contacts no current can pass, but as soon as a hole comes along the current gets through from roller to roller. The other system is on the principle of the Jacquard machine, which is used to control weaving looms; the perforations operate a number of small rods or 'needles.' In the telegraph transmitter it is the corresponding 'needles' which operate the electrical contacts. This indirect method of transmission has been found to be the most reliable.

The electrical current, which the perforations are to control, is made up of alternate positive and negative currents, as is also the case in the Wheatstone transmitter. In the Murray transmitter only one row of punched holes is used, whereas two rows are necessary in the Wheatstone. And, consequently, only one small upright rod for entering the perforations is required in the Murray, in place of two in the Wheatstone. The upper end of the small rod presses gently upwards against the underside of the paper-ribbon, while the lower end of the rod operates a combination of levers which make the desired electrical contacts. The transmitter is driven by a phonic-wheel motor, the electric impulses for which are timed by a vibrating reed or pendulum, the inertia of the phonic-wheel being sufficient to ensure uniform motion. This arrangement is very convenient, as variation of current strength has no effect upon the speed of the motor.

At the receiving station the telegraph current enters a distributing mechanism. The short electric impulses as they arrive are sent alternately to a punching-magnet and a spacing-magnet. When there are no signals arriving, a special device cuts out the spacing-magnet by opening a switch, and so stops the forward motion of the paper-ribbon. But as long as signals are passing, the vibratory action of the spacing-magnet is continuous, and the ribbon is fed forward at a definite rate. As already stated, the punching-magnet works alternately with the spacing-magnet, but only when there is marking current in the telegraph line.

The little punch in the receiving instrument is operated by the movement of the armature of the electro-magnet (punching-magnet). The movements of the armature are in turn under the control of the telegraph current, and that again is controlled by the movements of the small vertical rod, which is controlled by the perforations in the transmitting ribbon. In this manner the little punch in the receiver duplicates the movements of the small rod in the transmitter, the result being an exact duplicate of the perforated paper-ribbon. In the transmitter a perforation moves the rod, and in the receiver the moving punch makes a perforation.

The next step is to translate the perforations into Roman letters, by means of the automatic typewriter. The perforated ribbon passes over a drum, which operates in exactly the same manner as in a Jacquard loom. At each stroke of the machine the perforated ribbon is fed forward one letter-space, and then the drum moves forward against the points of five little rods or 'needles,' and then retires, to repeat the operation after having fed the ribbon along one more letter-space. But for the intervention of the paper-ribbon, the five needles could enter holes in the advancing drum, in which case they would remain inoperative. But whenever a solid piece of the paper comes opposite a needle, that needle is pushed back by the advancing drum; only a perforation in the paper allows the needle to remain as it was. It will be understood that all five needles take part in each stroke of the machine, but only those opposite which there is no perforation make any move. This part of the instrument is called the selector. The five rods are fixed in the ends of five slotted bars or combs, and no perforation means that the corresponding comb is pushed back about one-sixteenth of an inch.

The five combs are thin flat bars of steel, each having a series of slots, which form spaces between a row of square teeth along one side. They are not unlike the wards of a key continued in a long row. The five combs lie in a horizontal position, the one above the other, with their rows of teeth all facing a set of upright levers, which we shall call 'uprights.' Each upright is attached to a spiral spring which tends to pull the upright against the combs. We picture a regiment of fifty-six uprights, and in each comb there is a corresponding row of fifty-six slots or spaces between the teeth or wards. But when the combs are in their position of rest, the fifty-six spaces of one are not opposite the fifty-six spaces of the next one. Indeed, opposite every upright there is at least one tooth blocking the way into the combined terrace of slots. Therefore, when no needle is pushed back, no upright can be pulled into the slot belonging to it. Before any particular upright can be pulled forward by its spring, the five combs must each show a clear space immediately in front of the upright.

In front of one of the uprights we find that the way is clear except that one tooth on the uppermost comb blocks the way. If we slide this comb along one-sixteenth of an inch it moves the obstructing tooth out of the way and the upright falls forward. The comb will slide along in this manner when its needle is pushed back by the solid paper-ribbon on the advancing drum. If a perforation should be opposite that needle the upright will not be released. Opposite another of the uprights there is a clear space excepting in the second comb, so that no perforation opposite the second needle will operate this upright. We see that five of the uprights may be operated in this simple fashion. But another of the uprights has its way blocked by a tooth in each of two combs. In this case the corresponding two needles must be pushed back simultaneously. And so on we may go till we find an upright whose way is blocked by a tooth in each of four combs, which upright will require the corresponding four needles in the selector to be pushed back simultaneously. This movement will be obtained by having only one perforation opposite the needle of the comb whose space is already opposite the upright.

So far we see how each different signal or combination of perforations operates a different upright. When an upright is free to be pulled forward by its spiral spring, its upper end pushes forward a small hooked lever, which is hanging immediately in front of the upright. A metal bar or 'knife' descends at each stroke of the machine and will catch this hooked lever and pull it down. The hooked lever is hanging from one of the keys of the typewriter, so that the key is depressed and a letter printed on the page of paper, which is in position in the typewriter. When the knife rises and frees the engaged hook, the upright would still remain forward, under the tension of its spiral spring. However, at this moment a straight bar moves back against the upright and pushes it back till it is clear of the combs. This unlocking bar extends along the machine in front of all the uprights, and no matter which upright has been called for by the selector, that upright will be replaced in its normal position as soon as it has completed its task. All the uprights are thus held clear of the combs until the combs are set in position for the succeeding letter.

In this way the whole fifty-six keys of the typewriter may be operated individually, but it must be made possible to change the printer from letters to figures when necessary. This is accomplished by the addition of a sixth comb placed beneath the other five. When this sixth comb is in one position the machine is set to print letters, and when the comb is pushed along one-sixteenth of an inch the machine is then set for printing figures. The movement of this comb is obtained by two cross-bars, one at either end of the comb. The signal for the letter-shift key is the whole of the five holes punched in the paper-ribbon, thus calling for no letter-key. In this case, when none of the needles are pushed back by the perforated ribbon, the right-hand cross-bar is free to move forward into the slots of the combs. In the sixth comb there is a V-shaped slot, with one straight and one inclined face, while on the cross-bar there is a wedge-shaped piece with one inclined face. When the two inclined faces meet one another, the wedge pushes the comb along one-sixteenth of an inch, and the machine is set for printing letters. If it is desired to print figures, the cross-bar at the left-hand side is operated by a particular combination of punched holes (two perforations, one blank, two perforations). This cross-bar operates the sixth comb, just as was done by the other cross-bar, but moving it along in the opposite direction. In this position the machine will print figures.

During the operation of these cross-bars, it will be observed, there are no letter-keys called for. Hence by making a series of repetitions of five holes in the paper-ribbon, the printer will cease to act at each stroke. It is by this means that 'rubbing out' is attained; while the obliterating signals, referred to in an earlier paragraph, are passing through the printer it is quite inactive, and it is quite impossible to tell from the printed telegram whether or not the operator at the transmitting station made any corrections.


Many inventors have worked at the problem of seeing at a distance by means of electrical transmission. The problem is not solved, but its solution has commenced, and it may be that before the end of the twentieth century television will be a practical success.

We have seen how it is possible to transmit photographs by an electric current, but to transmit instantaneously a visual image is a much more difficult task. We wish to be able to throw the image of a moving person upon a screen which will transmit an electric current along a telegraph wire and reproduce the image at the receiving station.

Ernst Ruhmer, of Berlin, whose photo-grapho-phone is described in the succeeding chapter, gave a demonstration of his television apparatus in 1909. It was merely to show the principle of a much more elaborate system, which was estimated to cost about 50,000. The demonstration apparatus, by reason of its elementary construction, was capable of reproducing only a simple pattern of squares arranged in different combinations.

At the transmitting station an image of the pattern is thrown upon a screen hung upon a wall. The screen is divided into twenty-five square sections, and behind each of these is a highly sensitive selenium cell. At the receiving station there is a screen, divided into a similar number of square sections, each being connected by wire to the corresponding section in the transmitting screen. Each section controls a separate electric current, which is conducted to the corresponding section in the receiver. By a process similar in principle to the well-known mirror-galvanometer, the fluctuations of the current are made visible by a corresponding variation of light upon the receiving section.

Whatever pattern of dark and light squares is projected on the transmitting screen, a duplicate pattern appears simultaneously upon the receiving screen. The relationship of this simple pattern to the desired image of some natural object, may be realised by considering an analogy. In order to explain how photographs are reproduced in books by half-tone process blocks, it is convenient to magnify a small portion of the printed illustration. We see in the magnified part a collection of black dots crowded together in one portion and almost absent from another portion. In the magnified copy they do not form any resemblance to a portion of a photograph, but if viewed from some distance from which the individual dots cannot be seen, the effect is just such as we do see in looking directly at the printed illustrations, in which the individual dots are so small that they cannot be seen. We may take Ruhmer's simple screen as analogous to a greatly magnified portion of his would-be picture. It would require about 10,000 miniature sections to produce a complete television apparatus.

Two French inventors (Fournier and Rignoux) have devised an apparatus in which the transmitting screen is similar to Ruhmer' s, but with a receiving screen embodying totally different principles.

It should be mentioned that although the two transmitting screens are on exactly the same principle, the German and French inventors were working quite independently.

The first proposal by the French inventors happened to have a receiving screen also on somewhat similar lines to Ruhmer's invention. At the receiving station the incoming currents passed through little coils, or miniature galvanometers, which uncovered a corresponding number of little mirrors. The uncovering was in proportion to the amount of current received, which again was dependent upon the amount of light falling upon the controlling section at the transmitting station. At the receiving station, light is thrown upon the collection of mirrors, which reflect the light according to the amount of mirror space exposed. In this way, the receiving screen duplicates the image on the transmitting screen.

If we are to have 10,000 mirrors in either of these systems, it will mean as many individual connecting wires between the transmitting and receiving stations. But the French inventors believe that they have found a means of dispensing with this multiplicity of connecting wires, and it is in this system that they invented a receiver distinctly different from that in Ruhmer's invention.

In the demonstration apparatus they use a simple transmitting screen having only eight sections and eight selenium cells. The electric currents passing through the selenium cells are necessarily very weak, but they operate local relays which control more powerful currents, and it is these which pass out to the telegraph line. The eight different currents, from the eight cells, are to be conducted by a single telegraph wire to the receiving station. Each current is as it were to have the use of the line for a small fraction of a second, one current impulse following at the heels of the current impulse sent by the neighbouring section. The eight wires are connected to eight contact pieces in a circular switch called the collector. A rapidly rotating wheel carries a collecting arm which sweeps over the contacts in quick succession. This collector arm is connected to the telegraph line, and in this way the eight different currents are conveyed separately along the single telegraph line. It is evident that with a very rapid rotation of the collector, the time between the successive impulses from the same selenium cell will be only a very small fraction of a second. It remains to sort out these impulses at the receiving end.

The receiver in this single line system involves a principle which was discovered by Michael Faraday in 1845. Faraday discovered a direct connection between Light and Electricity. He polarized a beam of light by passing it through a Nicol prism (which permits only waves in one plane to pass), and then he passed the polarized beam through some transparent substance placed between the poles of a powerful electro-magnet. When the beam of light emerged from the magnetic field, it had to pass through a second Nicol prism, and thence to a screen upon which it would form a spot of light. This it could do only so long as the second prism was turned in the same position as the first prism. With the magnet disconnected from its source of current, the beam of polarized light was allowed to pass through the two prisms, and then the second prism was rotated about a quarter-turn until the beam could not pass through, and consequently the spot of light disappeared from the screen. If the electric current was then switched on to the magnet, the spot of light reappeared immediately. The cause of this phenomenon does not concern our present object; we have merely to deal with the facts.

In the receiver of this television apparatus, the old experiment of Faraday's is repeated. There is a lamp sending a beam of light through a Nicol prism, then through a tube containing carbon disulphide (a transparent liquid), and finally out through a second Nicol prism. The incoming telegraph line is connected to a coil of wire forming an electro-magnet around the tube. The arrangement of the prisms is such that no light passes until the magnetic field is present. The stronger the magnetic field is the greater will be the amount of light emerging from the second prism. The magnetic field will be dependent upon the amount of current received, and that again will be dependent upon the amount of light falling upon the selenium cells in the transmitting screen.

So far we have an ever-varying beam of light under the successive control of eight different selenium cells, the impulses following one another in very rapid succession. The next problem was to sort out the eight different currents, and piece them together again. But as the variations of the currents have been already translated into variations in the beam of polarized light, it will be more convenient to sort out the different variations of light corresponding to the different selenium cells which are producing them. This is done by allowing the beam of light to fall upon the periphery of a rotating wheel, upon which a series of small mirrors are fixed. There are as many mirrors on this receiving wheel as there are selenium cells in the transmitter. The wheel rotates in synchronism with the collector wheel in the transmitter, so that the beam while momentarily under the control of No. 1 selenium cell falls on No. 1 mirror, and so on. The little mirrors are so placed that each when receiving the beam of light reflects it to a different part of the receiving screen. For instance, in the simple demonstration apparatus, with its transmitting screen of eight sections, the part of the image falling upon the top left-hand corner of the screen will control the beam of light in the receiver at the moment when the beam is being reflected to the top left-hand corner of the receiving screen, and so on.

In this system, there will he still a multiplicity of wires at the transmitting station and a multiplicity of mirrors at the receiving station, if a complete image is to be transmitted. Neither of these conditions are impossible, and things have been very materially simplified by the abolition of the multiplicity of wires connecting the two distant stations together. A single wire is quite capable of carrying the individual impulses from, say, 10,000 separate selenium cells; the frequency of an alternating current may be 100,000 alternations per second.

To form the complete image of a natural object upon the receiving screen, the rotating wheel of mirrors will throw 10,000 patches of light upon the screen. These patches will be spread over the screen, just as the eight patches of the demonstration were. The patches will follow one another in quick succession, but the rapidity must be so great that, with the aid of our persistence of vision, all will stimulate the eye at one time. In the case of the kinematograph our persistence of vision enables us to see the quick succession of pictures on a screen as though there was no break in the continuity of the light.

It will be understood that no complete apparatus has been attempted, doubtless because of the great cost. But these simple demonstrations are of special interest, as they are the commencement of the solution of the problem of transmitting vision to a distance. We have become familiar with the transmission of sound to a distance, which was considered an impossible thing only forty years ago. Indeed, when Lord Kelvin told the members of the British Association about the telephone which he had seen at the Exhibition in Philadelphia (1876), and through which telephone he had heard extracts from New York papers, he said, "With my own ears I heard all this," and he assured them that there was no room for trickery, as it was his own friend Professor Watson, "who, at the other extremity of the line, uttered these words." It is true that the two problems, the reproduction of sound and the reproduction of vision, are not at all comparable, but there seems good hope for the ultimate solution of television.